DeepCode: A ‘Grammarly for programmers’ to clean up code..
Ask any computer programmer or software developer and one of the top work challenges they will cite is working on someone else’s code, as was also evident from the TechGig Developer Survey 2019.
Zurich-based tech platform DeepCode wants to change that and claims to have figured out a way to analyse and improve the code quality. There is no dearth of online tools for programmers out there, but the DeepCode tool has to be right on top of every developer’s list. The tool uses machine learning to automate code reviews and is being dubbed as the ‘Grammarly’ for programmers.
DeepCode will use 2.5 lakh rules, read your public and private GitHub repositories and tell you how to fix problems and improve your programs. DeepCode, the Swiss-based startup is founded by Veselin Raychev, Boris Paskalev, and Martin Vechev who have extensive experience in artificial intelligence (AI) and machine learning research.
Paskalev said, “We built a platform that understands the intent of the code. We autonomously understand millions of repositories and note the changes developers are making. Then we train our AI engine with those changes and can provide unique suggestions to every single line of code analysed by our platform.”
How does it work?
DeepCode reads your GitHub repositories and analyses them based on latest research in machine learning and other programming languages. The programming utility runs through many tools including JS Nice, Nice 2 Predict, and Deguard. It checks the code from over 250,000 coding rules, after that the built-in AI highlights problems associated with the code.